Socio-demographic patterns of posttraumatic stress disorder in Medellin, Colombia and the context of lifetime trauma exposure

ABSTRACT

Colombia, South America is currently transitioning to post-conflict status following 6 decades of armed conflict. The population has experienced extensive exposures to potentially traumatic events throughout the lifespan. Sources of trauma exposure include the prolonged armed insurgency, narco-trafficking violence, urban gang violence, violent actions of criminal bands, intra-familial violence, gender-based violence, and sex trafficking. Exposure to potentially traumatic events is related to a variety of psychiatric outcomes, in particular, posttraumatic stress disorder. Given this context of lifetime trauma exposure, socio-demographic patterns of posttraumatic stress disorder were explored in a sample of residents of Medellin, Colombia, the nation's second largest city and a nexus for multiple types of trauma exposure.

KEYWORDS:

• armed conflict
• Colombia
• posttraumatic stress disorder
• potentially traumatic events
• PTEs
• PTSD
• rurality
• trauma exposure
• urbanicity

Background

Throughout history, humans have faced trauma associated with armed conflict, community violence, environmental disasters, maltreatment and sexual abuse, and a variety of adverse experiences that carry significant negative psychological consequences. Exposure to “potentially traumatic events” (PTEs)¹ may profoundly compromise the biological, psychological, and social integrity of individuals, and their communities, leading some individuals to develop trauma-related psychiatric disorders.² The most recognized and researched psychiatric disorder associated with trauma exposure is posttraumatic stress disorder (PTSD).³ This paper explores socio-demographic patterns and correlates of PTSD in urban and rural zones of Medellin, the capital of the Colombian department (state) of Antioquia.

PTEs and PTSD in the United States

To provide context for the discussion of PTSD in Medellin, Colombia, the relationship between PTEs and PTSD is first examined by referencing the broader scientific literature. The preponderance of recently published studies that examine the socio-demographic correlates of PTSD come from the United States and other highly-developed nations. This paper begins with an overview of the international literature prior to shifting the focus to the experience of Medellin. The globally accepted diagnostic standard for PTSD is the American Psychiatric Association's Diagnostic and Statistical Manual fifth edition.⁴ The diagnosis of PTSD is based on the individual's direct exposure to trauma (Criterion A1) coupled with multiple symptoms drawn from 3 domains: 1) pervasive re-experiencing of features of the original traumatic event, 2) hyper-vigilance with autonomic activation, and 3) avoidance of situations or circumstances that serve as traumatic reminders of what happened.⁴

As described by Fink and Galea (2015), exposure to PTEs is ubiquitous worldwide and, understandably, lifetime prevalence increases with age.⁵ Indeed, in the Mental Health Surveillance Study (MHSS) that enrolled a representative sample of 5,653 United States participants, 40.8% reported lifetime exposure to one or more PTEs.¹ Kilpatrick's (2013) National Stressful Events Survey estimated lifetime exposure to a traumatic event to be much higher, 89.7% based on DSM-5 criteria and 93.7% based on the predecessor DSM-IV criteria.⁶

The counterpoint to these high prevalence rates is that most persons who have experienced PTEs do not progress to PTSD. Kilpatrick's team found lifetime PTSD prevalence to be 9.4% (DSM-5) or 10.6% (DSM-IV). The corresponding past-year PTSD estimates were 5.3% (DSM-5) and 6.9% (DSM-IV), respectively. Both of these figures were higher than the past-year PTSD estimate of 3.5% that Kessler et al. (2005) had previously reported based on data from the National Comorbidity Survey Replication in the United States.⁷

Severity of exposure is a qualifying feature when determining the occurrence of PTSD. As reported by North and Pfefferbaum (2013), following community disasters involving widespread and collective trauma, PTSD may be diagnosed in “one-third” and major depression in “one-fourth,” of intensely-exposed survivors.^8-13

Socio-demographic correlates of PTSD

Descriptive epidemiologic analyses conducted in developed nations have examined PTSD patterns in relation to a range of socio-demographic indicators. These descriptors include ethnicity,¹⁴ age,¹⁵ gender,¹⁶ race,¹⁷ and immigration status.¹⁸ PTSD rates have been estimated for various types of trauma and victimization (intra-familial and community violence; natural disasters; physical and sexual abuse).^19-21 PTSD rates have also been calculated in relation to the presence of comorbid conditions such as major depression and substance abuse or dependence.^22,23

Ethnicity

Studies that focused on the impact of ethnicity on the development of PTSD suggest that Latinos who have suffered trauma are at higher risk for endorsing PTSD symptoms than persons from other ethnicities who lived through similar experiences.¹⁴ One of the first reports to document this disparity was conducted by the National Vietnam Veterans Readjustment Study.²⁴ In this study, the PTSD rate for Hispanic male combat veterans was 7.3% higher than for their African American counterparts and 14.2% higher than for white, non-Hispanic veterans. Moreover, the symptoms Hispanic American combat veterans experienced seemed to be more severe than the symptoms suffered by non-Hispanic combat veterans.²⁵ The same phenomena regarding PTSD rates and symptom severity were observed among police officers during a survey conducted by Pole et al. in 2001.²⁶ Perilla, Norris and Lavizzo (2002)²⁷ found that Latino survivors of Hurricane Andrew had higher rates of storm-related PTSD than survivors from other cultural backgrounds. Furthermore, in the aftermath of September 11, a survey of New York residents revealed that 14% of Hispanics met criteria for PTSD in contrast to 9% of African Americans and 7% of white, non-Hispanic residents.^28,14

Different explanations have been offered to clarify why Latinos may be more likely to develop PTSD than individuals from other ethnicities. It has been hypothesized that Latinos' tendency to believe in destiny or “God's will” makes them less likely to seek help, due to the idea that there is nothing they can do to change their situation. Hispanics' belief that they cannot alter their destiny, that they are bound by fate, is called fatalismo. This belief in fate as a determinant of the life course can lead Hispanics to take a passive attitude that can be detrimental for their recovery.¹⁴

Additionally, studies have shown that Latinos tend to express their emotional distress through physical symptoms, a characteristic that may increase the likelihood that Latinos match the criteria for PTSD in the aftermath of trauma exposure. A report from the World Health Organization (WHO) revealed that Latinos in South America were 10 times more likely to suffer from somatization disorder than citizens from non-Hispanic countries.²⁹ Research conducted in Colombia found comparable high rates of somatization disorder in samples drawn from the populations of communities with or without intensive exposure to armed conflict (73% vs. 61%).³⁰ The incidence of somatic symptoms seems to be related to Hispanics' tendency to see mind, body, and spirit as a continuum and, importantly, individuals with somatization disorder are 3 times more likely to develop PTSD.¹⁴

Furthermore, Latinos tend to experience peri-traumatic dissociation or “altered states of consciousness” during and following traumatic exposure.¹⁴ This tendency is one of the best predictors of PTSD.¹⁴ Alejo et al. (2007)³¹ conducted a study on internally displaced adults in Colombia and found that peri-traumatic dissociation was one of the strongest risk factors for the development of PTSD symptoms.

Urbanicity

Studies have examined rates of mental health disorders in relation to urbanicity.^32,38 Peen and colleagues (2010) conducted a large-scale meta-analysis of international studies across 16 developed nations, spanning 25 y.³⁴ Pooled total prevalence rates for psychiatric disorders were significantly higher for urban settings compared with rural areas, a finding that was consistent across total disorders, mood disorders, and anxiety disorders. As an important caveat, only one of the articles included in this meta-analysis discussed PTSD specifically and this paper did not provide sufficient evidence to support the notion of higher PTSD rates in urban areas.

A series of studies conducted in the United States examined patterns of mental disorders in relation to “urbanicity” or “rurality.” Probst et al. (2006)³⁹ applied the Composite International Diagnostic Interview Short Form (CIDI-SF) depression scale to a nationally representative sample of 30,801 adults, ages 18 and over. These investigators found that adjusted rates of major depression did not differ significantly for urban versus rural samples. A 2014 study by Breslau et al.⁴⁰ used 4 categories of urbanicity (large metropolitan, small metropolitan, semi-rural, and rural) and observed that rates of major depression did not differ between large metropolitan and rural areas. However, there was a small but significantly higher rate of major depression in the intermediate categories (small metropolitan and semi-rural).

Specific to PTSD rates in the United States, Reeves and colleagues (2013)⁴¹ compared metropolitan Atlanta to urban (moderate size metropolitan areas) and rural areas in the state of Georgia. PTSD rates were significantly higher for both urban and rural areas compared with the Atlanta metropolitan area. McCall-Hosenfeld and her research team (2014)³² used rural-urban continuum codes (RUCCs), a 9-point classification ranging from 1 (most urban) to 9 (most rural), to examine for possible differences in the prevalence of mental disorders in the National Comorbidity Survey Replication (NCS-R) by “rurality.” Substance use was higher in metropolitan areas as was war-related trauma. However there were no differences in rates of PTSD across metropolitan, urban, and rural areas. Studies of United States combat veterans have found higher rates of PTSD in rural veterans⁴² and in Latino veterans, especially those with rural residence.³⁷

PTSD studies from Medellin, Colombia

Trauma exposure in the Colombian context

Colombia, South America is distinguished by widespread population exposure to PTEs in a manner that affects both individual and collective mental health.⁴³ Colombia has been the scene of one of the longest continuous armed conflicts, pitting left-wing guerrilla and right-wing paramilitary forces against the Colombian military and National Police, resulting in an estimated 250,000 deaths. Throughout the nation there is nearly-universal exposure to trauma and loss related to the protracted armed insurgency, narco-trafficking (Colombia is the world's leading source nation for cocaine and one of the 2 primary source nations for heroin entering the United States), actions of criminal bands, community violence, intra-familial violence, and sexual and gender-based violence including sex trafficking.^43-45

The focus on Medellin is appropriate for examining this theme because this area of Colombia has been a nexus for multiple streams of violence. The department of Antioquia has been a geographic power base for multiple armed actors and the epicenter of operations for the notorious Medellin drug cartel. Large proportions of the rural population surrounding Medellin were uprooted and internally displaced, while the urban core would ultimately become one of Colombia's hub cities for relocation of these forced migrants (a “receptor” city).

In broader context, Medellin is the second-largest city in Colombia, South America with an estimated population of 2.44 million citizens. Situated in the Central Range of the Andes, Medellin is the capital of the department of Antioquia. Medellin is notable for its universities and academic institutions, health services, banking and commerce, and in the rural environs, and cultivation of flowers for global export. In February 2013, the Urban Land Institute chose Medellin as the most innovative city in the world due to its recent advances in politics, education, and social development.

Data analysis approach

Psychiatrists and researchers from Universidad CES in Medellin performed secondary analyses on data from the First Population Mental Health Study of Medellin, Colombia, a cross-sectional observational study that was conducted in 2011–2012.⁴⁶ The study applied the WHO Composite International Diagnostic Interview (WHO-CIDI) (v3.0), validated in Spain, as the data collection instrument.^46,47 This is a Spanish language version of the World Mental Health (WMH) Survey that has been used by Harvard University and WHO in their renowned transcultural psychiatric epidemiological study comparing 28 nations.^48,49 The WHO-CIDI consists of 2 parts; the first part identifies subjects at risk for any type of mental disorder and the second estimates the prevalence of 21 mental disorders based on DSM IV criteria.⁵⁰

Universidad CES researchers enrolled a systematic random sample of 4,176 citizens from Medellin, Colombia, and among these, 3,264 participants were within the age range 19-65 y. For these age-eligible adults, 1,143 screened positive for the presence of a mental disorder and all were entered into the secondary analyses. An additional 530 subjects were added, representing a 25% random sample of the remaining 2,121 survey participants who screened negative for any mental disorder, bringing the total group selected for secondary analyses to 1,673 subjects.

As an important caveat, this selection process– entering 100% of subjects with possible mental disorders (screen positives) but only 25% of those without disorders (screen negatives) – may have distorted the demographic makeup of the analyzed sample and surely increased the prevalence rate estimates for PTSD and other mental disorders when compared with the original representative sample. In the original sample, the prevalence of screen positives was 35.0% (1,143/3,264). In the reconstituted secondary analysis sample that eliminated 75% of screen negatives, the prevalence of screen positives was 68.3% (1,143/1,673). Thus, the prevalence of probable mental disorders was roughly double that of the original population-based sample. Also unclear is how the removal of 75% of the screen negatives affected the demographic character of the analyzed sample for important measures such as gender and urbanicity.

Descriptive statistics for the selected study group focused on a range of socio-demographic variables including age, gender, marital status, educational level, and rural/urban residence. Regarding mental health disorders, the presence or absence of PTSD was the independent variable. The presence of major depression, dysthymia, bipolar disorder I and II, alcohol dependence, and drug dependence was also assessed. The primary data analyses compared urban vs. rural Medellin residents, with or without PTSD, in relation to the other socio-demographic and clinical variables.

Demographics and PTSD patterns (Table 1)

Among the 1,673 adults analyzed, 1,306 (78.1%) were classified as urban residents and 367 (21.9%) were rural residents. The mean ages were comparable: 42.3 y for urban participants and 43.8 y for rural interviewees.

Table 1. Posttraumatic stress disorder prevalence for urban and rural residents of Medellin, Colombia by gender and age category.

	Urban			Rural			Total
	PTSD	n	%	PTSD	n	%	PTSD	n	%
All participants	79	1,306	6.05	30	367	8.17	109	1,673	6.52
Gender
Male	11	430	2.56	6	101	5.94	17	531	3.20
Female	68	876	7.76	24	266	9.02	92	1142	8.06
Age groups
19-29 years	16	314	5.10	4	95	4.21	20	409	4.89
40-44 years	27	430	6.28	11	131	8.40	38	561	6.77
45-65 years	36	562	6.41	15	141	10.64	51	703	7.25
Marital status
Single	23	511	4.50	6	105	5.71	29	616	4.71
Cohabiting	23	238	9.66	8	86	9.30	31	324	9.57
Married	15	352	4.26	9	121	7.44	24	473	5.07
Separated/Divorced	10	148	6.76	5	34	14.71	15	182	8.24
Widowed	8	57	14.04	2	21	9.52	10	78	12.82
Highest educational attainment
Less than primary	1	26	3.85	2	17	11.76	3	43	6.98
Primary	20	257	7.78	14	135	10.37	34	392	8.67
Secondary	36	549	6.56	4	128	3.13	40	677	5.91
Vocational/Technical	16	221	7.24	7	70	10.00	23	291	7.90
University	6	202	2.97	3	11	27.27	9	213	4.23
Postgraduate	0	51	0.00	0	6	0.00	0	57	0.00

Note. The sample is based on 100% of screen positives and 25% of screen negatives from a representative population sample. Therefore the percentages tend to overestimate the true prevalence rates.

A total of 109 of 1,673 participants were assessed as meeting criteria for PTSD, representing a lifetime PTSD prevalence estimate of 6.52%. Ninety-two (84.4%) were women and 17 (15.6%) were men. This subset of 109 participants included 79 urban residents (PTSD prevalence: 6.05%) and 30 rural residents (PTSD prevalence: 8.17%). The mean age of PTSD onset was 22.4 y for 79 urban participants and 25.7 y for 30 rural interviewees meeting PTSD criteria.

The rate of PTSD was significantly higher for women (92 of 1,142 female participants met PTSD criteria, 8.06%) than for men (17 of 531 male participants, 3.20%). When PTSD prevalence was analyzed jointly according to gender and place of residence, the combined effects of higher PTSD rates for women and persons living in rural zones created the following rank ordering of PTSD prevalence rates: male urban, 2.6%; male rural, 5.9%, female urban, 7.8%, and female rural, 9.0%. Furthermore, the gender disparity was greater for urban participants (3 times higher rate for women compared with men) than for rural residents (1.5 times higher rate for women).

Rates of PTSD tended to increase with age group for both urban and rural residents, with a more pronounced upward progression in prevalence for the rural citizens. Lower educational level and higher rates of illiteracy were found in rural areas compared with urban zones.

Psychiatric and substance abuse/dependence comorbidity

Major depression was co-occurring for 58% of participants meeting criteria for PTSD, with similar rates of comorbidity for persons from urban and rural zones. About 21% of participants had concurrent substance use disorder, a category that subsumes alcohol and drug abuse and dependence; again rates of comorbidity were comparable for the urban and rural interviewees.

Bivariate and multivariate analyses

The dependent variable, PTSD, was compared with all the other variables included in the study. Female gender was the dominant demographic predictor of PTSD with an odds ratio (OR) of 2.65 (p = 0.0002). Among clinical variables, major depression displayed the strongest relationship to PTSD of any variable (OR: 6.12, p < 0.0001). Alcohol dependence (OR: 2.64, p = 0.0023) was also significantly related to PTSD. Both major depression and alcohol dependence continued to be significant predictors when examined separately for urban and rural residents. Although drug dependence showed a particularly strong relationship to PTSD among rural residents, it was not related to PTSD for urban participants. For the urban zone analysis, female gender, major depression, and alcohol dependence were retained in the model; for the rural zone analysis, only major depression was predictive of PTSD.

Summary of main findings

A secondary analysis was conducted using data on persons, ages 19–65 years, from the First Population Mental Health Study in Medellin, a large cross-sectional study conducted with a representative sample of urban and rural residents. Socio-demographic characteristics and psychiatric/substance use comorbidities were also examined in the analysis.

Five primary findings are noted. First, the overall lifetime PTSD prevalence rate, 6.5% is almost certainly an overestimate due to the sampling procedures that enrolled all screen positives but only 25% of screen negatives. Nevertheless, even this inflated figure is considerably lower than Kilpatrick et al.'s⁶ 10.6% rate for the United States (using the DSM-IV criteria as used in Medellin).

Second, consistent with findings throughout developed and developing nations, the PTSD rate in women was higher than for men. Third, regarding urban/rural comparisons, the PTSD rate in rural zones was about 35% higher than the rate in urban sectors of Medellin. Fourth, when examining gender and urban/rural residence simultaneously, PTSD prevalence rates were highest for rural females and lowest for urban males. Fifth, an estimated 58% of persons with PTSD had comorbid major depression, with nearly identical proportions for urban and rural residents. Alcohol dependence was a common co-occurring disorder.

Significance of findings

The higher PTSD prevalence rate in women, the frequent co-occurrence of major depression with PTSD, and the significant contribution of alcohol dependence, are consistent with findings from mental health surveys globally. It is noteworthy that the study's finding of a 6.5% lifetime PTSD prevalence rate in a sample that had twice the proportion of screen positives as the general population, was nevertheless lower than baseline PTSD rates in the United States. These findings may reflect reluctance on the part of study participants to disclose psychiatric symptoms due to stigma, differential responses following trauma exposure, or the operation of resilience factors that may have developed in Colombian citizens following decades of frequent and repetitive trauma exposure. The finding of a higher rate of PTSD in rural dwellers compared with their urban counterparts may reflect disparities in the frequency and intensity of exposures to trauma, loss, and change, as will be discussed below.

Discussion

Urban/rural disparity in PTSD prevalence

An important caution to highlight is the fact that the rural sample (n = 367) was much smaller than the urban sample (n = 1,306) and the total number of rural respondents meeting criteria for PTSD was only 30 persons, including just 6 rural males. These small cell sizes limit the ability to confidently assert that rural PTSD prevalence rates are higher than urban rates. However, in support of this notion, PTSD rates were higher in rural settings for both men and women and the gender disparity was less in the rural settings. Also, in the multivariate model, only depression was significantly related to PTSD, suggesting the need to determine whether higher rates of psychopathology (both depression and PTSD) in rural zones may be related to differential exposures to PTEs.

Exposure to armed conflict and violence

Numerous studies have found that exposure to armed conflict negatively affects the mental health of civilians. High rates of psychiatric symptoms have been observed among individuals impacted by conflict in diverse settings worldwide, including Lebanon,⁵¹ Cambodia,⁵² Afghanistan,⁵³ the Balkans,⁵⁴ Rwanda,⁵⁵ Algeria, Gaza, and Ethiopia.⁵⁶ Similar findings derived from studies in Latin American countries impacted by conflict and violence. Pineda et al. (2002)⁵⁸ conducted a study in a Colombian village that was attacked by guerrilla forces 2 y prior to the study and found that 37.6% of respondents met PTSD criteria. A significant association was observed between exposure to armed conflict and the prevalence of psychiatric disorders while comparing the citizens of Guasca, a Colombian “pueblo” (small town) that had been occupied by the guerrilla and the citizens of Guatavita, a nearby Colombian town where armed conflict never occurred.³⁰ Bell et al. (2012) recently conducted a study in 4 Colombian provinces that had experienced high levels of violence related to the armed conflict. Investigators created a three-category typology of violence: armed-conflict violence, domestic violence, and a residual category of “other” violence. Conflict-related violence was associated with elevated levels of fear, anxiety, and hyper-alertness. Non-conflict-related violence was associated with aggression and a higher prevalence of substance use. Prevalence rates for depression, including suicidal thought or intent, were similar across the three categories of violence.

Forced displacement and PTSD according to place of residence

An important factor in this analysis is conflict-induced internal displacement in Colombia. Internally displaced persons (IDPs) are those who have been forced to abandon their homes or habitual places of residence to avoid the threats and harmful effects of armed conflicts but who continue to reside within the borders of their country of origin. Colombia has ranked first or second worldwide in numbers of IDPs for 12 consecutive years, and in 2015, the estimate for total Colombian IDPs was 6.3 million. This includes more than 600,000 IDPs in the department of Antioquia where Medellín is the capital.⁶⁰ IDPs comprise 10.4% of the Antioquia's population, the highest rate of any of Colombia's 32 departments. Caceres et al. (2002) ⁶¹ found that 80% of displaced adults in Colombia reported exposure to violence before displacement. The prevalence of PTSD among different groups of adult IDPs in Colombia has been estimated at 21% in Bucaramanga,³¹ 29% in Barrancabermeja,⁶² and 88% in Medellin.⁶³

Internal displacement may be a factor that contributes to differential PTSD rates. Rural Antioquia, including rural areas in proximity to Medellin, are principally areas of “expulsion” while urban Medellin represents a sanctuary for “reception” and relocation of IDPs. Also of note, 25% of people residing in rural zones are widows and widowers, compared with 14% in the urban zones.

Depression and PTSD

Major depression is a common co-morbidity for persons diagnosed with PTSD,^59,63,64 The co-occurrence of PTSD and depression is associated with significantly higher subjective distress and functional impairment than PTSD diagnosis in isolation.^65,66 A meta-analysis conducted by Rytwinski et al.⁶⁷ found that the average co-occurrence of depression in people with PTSD was 52%, and that depression is 4 times more frequent in individuals already diagnosed with PTSD, regardless of place of residence. In the Medellin analyses, major depression was the only variable that showed a significant predictive relationship to PTSD.

The same authors^67,68 found that gender was not significantly connected to the co-occurrence of depression and PTSD. This may also have clinical significance because it is known that irritability and hostility are frequent depressive symptoms, particularly in men, constituting a clinical picture that could be confused with PTSD,⁶⁹ thus making it harder to arrive at a valid diagnosis of depression. The situation becomes even more complicated because male patients are generally more reluctant to seek help when experiencing emotional symptoms.⁷⁰ Ritwinsky and colleagues also found that in rural areas, the onset of depression coincides with, or occurs after, PTSD more often than in urban zones.⁶⁷

Gender

In the Medellin study, PTSD prevalence was higher for women than men, consistent with international findings. Gender differences in mental health responses have been found in populations exposed to multiple types of trauma, including those impacted by displacement and armed conflict.^61,71-75 Different neural mechanisms in men and women have been proposed to explain these gender disparities.^76,77 In women, higher sensitivity to a corticotropin-releasing factor has been reported,⁷⁸ potentially producing higher rates of depression, peritraumatic fear, dissociation, and negative peritraumatic cognitions, in comparison to men's responses.^79,80 On the other hand, men who have experienced trauma have been found to exhibit externalizing behavioral disorders such as substance abuse and somatic symptoms, more frequently than their female counterparts.^7,81

Gender-based violence

PTSD among women in Colombia could also be partially explained by high rates of exposure to gender-based violence. In 2011, the National Institute for Legal Medicine and Forensic Sciences of Colombia (Instituto Nacional de Medicina Legal y Ciencias Forenses de Colombia) reported 1,371 victims of sex crimes, of which 88.8 % were women.⁸² Regarding domestic violence, 5,649 victims were reported throughout Colombia in 2011 and 82.6% were women. In the 2010 National Demographic and Health Poll–ENDS (Encuesta Nacional de Demografía y Salud),⁸³ 37% of the women reported that they had experienced physical violence inflicted by their partners, and 6% reported that they had been victims of sexual violence. Perrin et al.⁸⁴ found that the risk factor most frequently related to PTSD in the overall population was sexual abuse, followed by bipolar disorder, alcohol dependence, antisocial personality, and having been the victim of a criminal action.

Alcohol and PTSD

The Medellin secondary data analysis found a close relationship between alcohol dependence and PTSD in urban and rural zones, yet the figures were higher in rural zones. A bidirectional relationship appears to exist between PTSD and alcohol dependence. In the study, 28% of PTSD patients were alcohol-dependent and high proportions of male and female participants with alcohol dependence met criteria for PTSD, a finding consistent with other published research.^85-88

In Colombia, as elsewhere, men drink more frequently and in larger amounts than women.^89,90 In a Colombian national epidemiological study, Posada-Villa et al (2008)⁹¹ found significantly higher lifetime alcohol abuse prevalence rates in men compared with women (13.2% vs. 1.6%). Since consuming alcohol is considered socially acceptable in Colombia,⁹² it has been hypothesized that Colombian men with mental health issues may use alcohol to cope with their symptoms instead of reporting distress.⁷³

Study limitations

Several important study limitations must be acknowledged. First, the sampling approach limited the ability to develop robust prevalence estimates due to the non-inclusion of 75% of the screen negatives, as previously discussed. Second, as a cross-sectional study, it is not possible to establish cause-and-effect relationships. Third, participants were not asked about their status as internally displaced persons or other categories of “victims of the armed conflict.” Fourth, several PTSD-related mental disorders such as anxiety and personality disorders were not assessed. Fifth, the study did not assess trauma exposure generally or ask about specific traumas sustained prior to PTSD diagnosis. This limits the ability to directly ascertain the relationship of PTSD to armed conflict or various forms of narco-trafficking, community violence, or interpersonal violence. Sixth, no information was gathered regarding history of psychiatric treatment.

Despite these limitations, the present study contributes to the knowledge base regarding socio-demographic and clinical variables in relation to PTSD prevalence in Colombia.

Conclusions

The examination of socio-epidemiological determinants of psychopathology, and specifically PTSD, is a fruitful area for research. The selection of Medellin as a site for studying gender, urban/rural residence, and other socio-demographic contributors to PTSD prevalence is important because of the intensity and duration of exposure to PTEs in this population across 6 decades of continuous armed conflict.

Follow-up studies are encouraged to better characterize the PTEs and the co-occurring mental disorders in relation to PTSD incidence and prevalence. The urban/rural comparisons in particular will benefit from larger studies that carefully characterize the current and historical trauma exposures experienced by this population. Such research also provides the underpinnings for adapting effective, evidence-based interventions to address the population burden of PTSD, major depression, and alcohol abuse and dependence.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Acknowledgments

The authors wish to thank the Center for Excellence on Research in Mental Health, Medellín (Colombia) and the Secretary of Health, for access to the data and assistance with instrumentation, fieldwork, and consultation.

Funding

The research reported here was supported in part by US Public Health Service Grant 1 R01MH101227–01A1 and was performed in conjunction with WHO's WMH Survey Initiative, supported by the National Institute of Mental Health (R01 MH070884 and R01 MH093612–01), the John D. and Catherine T. MacArthur Foundation, the Pfizer Foundation, the US. Public Health Service (R13-MH066849, R01-MH069864, and R01 DA016558), the Fogarty International Center (FIRCA R03-TW006481), the Pan American Health Organization, Eli Lilly and Company, Ortho-McNeil Pharmaceutical, GlaxoSmithKline, and Bristol-Myers Squibb. None of the funders had any role in the design, analysis, interpretation of results, or preparation of this paper. The views and opinions expressed in this report are those of the authors, and should not be construed to represent the views of the sponsoring organizations, agencies, or governments.